Warning: Don’t ever put a UV light on an undersink reverse osmosis system!

I have to admit that we used to sell an under-sink reverse osmosis system that included an ultraviolet or UV light, but if you look on our site now, you will not find anything of that type.  Many of our competitors do have such systems, however.  We have decided that we will no longer sell an undersink UV for a reverse osmosis system.  Are we missing the boat?  Well, I am going to present you with the facts and you decide.

First of all, why do you want an ultraviolet light on your reverse osmosis system?  Likely you will say that you want UV so that you can make sure your water is pure and free of bacteria.  A reverse osmosis system filters down to below .001 microns, but the smallest bacteria is larger than .01 microns, so theoretically, a RO system should not allow any bacteria to pass through.  However, if you look at almost any reverse osmosis system manufactured by any credible manufacturer, you will find that it says “Not for use on water that is not microbiologically pure.”  This is because a RO system has mechanical seals and o-rings that could become compromised or fail, so you should not rely on reverse osmosis for bacterial purification.

Here’s one problem:  Even if your water is contaminated with bacteria, you still bathe, shower and brush your teeth in it.  A couple of years ago, I showered in contaminated water in Haiti and got infections in both eyes.  It wasn’t fun, I can tell you for a fact.  If you have children or grandchildren, you know that they drink from anywhere, even the bathtub.  An ultraviolet light on the RO system does nothing for the rest of the water in the home.  You would be better served with putting an ultraviolet light on the incoming water supply to treat the water before the reverse osmosis system, so that you are not installing the RO system on water that is microbiologically contaminated.

Here’s a second problem:  An ultraviolet light on a reverse osmosis system under your kitchen sink heats the water as it sits during the day or night.  It’s almost hot enough to make your coffee in the morning!  You could turn the UV light off, but then you could not be sure if your water was pure and turning the bulb on and off would shorten bulb by by up to 300% – not a good idea.

The third problem:  The heat generated by the UV light bulb can weaken fitting, embrittle plastic tubing and create a leak in certain circumstances.  That’s really pretty undesirable in most modern kitchens.

The forth problem:  To put a UV under your sink, you need electricity.  I have heard people say “I have electricity to my garbage disposal” and they do, but your UV system will only work when you turn your garbage disposal on (not a good idea).  You need to run an additional electric line to the ultraviolet system and that can often be a problem.

The final (fifth) problem:  Most undersink ultraviolet systems are cheap plastic Asian Import models or cheap, poorly designed stainless steel Asian Import models.  I’ll be kind and just say that they inherently have a lot of problems, not the least of which are leaks.

So, if you want a UV for use after your reverse osmosis system, the odds are, once you weigh the pros and cons, you will decide against it.  The best solution is to install a ultraviolet system on the point-of-entry (POE) into your home so that all of the water in your home is bacterially pure.

This article has 20 Comments

  1. Thank you for the info on U. V systems location. I use Severn River (Ontario Canada)water thru a UV and am installing a R.O. system and was about to mount the R.O. ahead of the U.V. as shown on most Water Purification Companies web sites. Thinking about it over night I said to my self, currently I’m killing the bacteria on all my incoming water, now if I change I’m only treating the drinking water not the rest. Your site has just verified what I was thinking and given me confidence to go ahead. I will have it tested of coarse. The UV heat issue is also very real and very interesting. My unit is in the basement, 20′ away. Andy

  2. You sell the Polaris UV-1C unit which appears to be positioned as an add-on to a reverse osmosis system. What application would that be used for if not undersink?

  3. Sorry to disagree with your conclusion. There is a reason to install the UV lamp AFTER the RO Membrane. Putting the UV unit before the Reverse Osmosis step has one large disadvantage: The UV lamp housing has a finite volume. Therefore, the RESIDENCE TIME that the water is under the UV rays is inversely proportional to flow in GPM. You need to realize that Reverse Osmosis membranes used in household water purification units require several times the water throughput in relation to the permeated water that is filtered and delivered by the membrane. In other words, for each gallon of pure water the membrane produces, the unit has to pass around about 7 to 9 times more water around the membrane to keep it clean and avoid plugging it. This waste stream is called “the Reject” (and is piped to the drain).
    By putting the UV lamp before the membrane, the effective flow that passes under the UV rays has to be several times the flow produced by the membrane. So, by installing the UV unit before the membrane, you end up reducing the effective time the water is irradiated with the UV rays, so its disinfection effect is cut to a fraction of the time it will get when put after the membrane. That is the reason home RO systems have the lamp after the membrane, on the permeated water outlet. Water heating by the UV lamp is because of the UV lamp having filaments that produce heat, but it is of no real consequence (other than warming the water for the first couple of glasses, which bothers most users but has no other detrimental consecuences, as the UV irradiated water no longer has enough live bacteria to cause damage.

    Installing an effective UV lamp at the Point-of-Entry requires a very large UV sterilizer device, as the total water flow at that point is many times the one after the RO membrane unit. Household RO systems that come with UV lamps have small wattage lamps, which is quite satisfactory for the small flow they produce. But pretending to correctly sterilize with UV light ALL the water at the point of entry is an entirely different matter. In order to be able to kill most of the bacteriae, the UV light intensity, wavelenght AND the irradiation time all need to be met. Unless you install (and can buy) a powerful enough UV lamp sterilizer, large enough to treat ALL the incomming water is not only difficult, but expensive, and wasteful too, as the RO membrane has to send the most of the water as the reject.
    Now, speaking of home RO systems with UV, good quality UV lamps (like those from Philips Brand) are the ones NEEDED in an effective unit, as gray market substitutes, even when rated at a higher wattage, do not really produce strong enough UV output in the correct wavelenght. So, check that a quality lamp is used in your unit and place it AFTER the membrane,in order to irradiate the water for a long enough time to kill bacteria. As a side note, I have installed RO units in Offshore Oil and gas Platforms, large enough to feed a 150 person hotel (about 13,200 gallons per day), and all of them use the UV irradiators AFTER the permeators or membranes. Best wishes. Amclaussen, Chemical Engineer.

  4. Alfredo,

    Sorry to disagree with your disagreement, but you are attempting to compare residential water treatment to commercial/industrial water treatment and you have drawn several wrong conclusions and made numerous mis-statements:

    You said: “The unit has to pass around about 7 to 9 times more water around the membrane to keep it clean and avoid plugging it.” Reality is, most residential RO’s waste 3 to 4 gallons of wter for every gallon made, not 7-9 gallons!

    You said: “By putting the UV lamp before the membrane, the effective flow that passes under the UV rays has to be several times the flow produced by the membrane. So, by installing the UV unit before the membrane, you end up reducing the effective time the water is irradiated with the UV rays, so its disinfection effect is cut to a fraction of the time it will get when put after the membrane.” Well, that’s just plain silly – put the right size UV on the system and you are fine. You can’t put the same size UV before an RO as you would after. That’s simply a non-issue.

    You said “[heating the water]is of no real consequence (other than warming the water for the first couple of glasses, which bothers most users but has no other detrimental consequences, as the UV irradiated water no longer has enough live bacteria to cause damage”. Reverse Osmosis systems use plastic fittings and tubing. This causes fitting to become embrittled and can have disastrous results with leaks under the sink. I’m sure that a homeowner who has experienced a “flood” in their kitchen will not agree that there are no detrimental consequences.

    You said: “Installing an effective UV lamp at the Point-of-Entry requires a very large UV sterilizer device, as the total water flow at that point is many times the one after the RO membrane unit. Household RO systems that come with UV lamps have small wattage lamps, which is quite satisfactory for the small flow they produce. But pretending to correctly sterilize with UV light ALL the water at the point of entry is an entirely different matter. In order to be able to kill most of the bacteriae, the UV light intensity, wavelenght AND the irradiation time all need to be met. Unless you install (and can buy) a powerful enough UV lamp sterilizer, large enough to treat ALL the incoming water is not only difficult, but expensive, and wasteful too, as the RO membrane has to send the most of the water as the reject.” If a UV is needed to kill bacteria, why would you only want it on your drinking water at the kitchen sink, when you brush you teeth in the bathroom (with contaminated water), you shower in contaminated water (have you ever got an eye infection from contaminated water? I have) and your kids bathe in contaminated water.

    We sell a UV for an RO that is about $100. You can protect the whole house for about $500. Why on earth would anyone think about just putting it on the RO? Besides (and this is the kicker), a reverse osmosis system should not be put on water that is not micro-biologically pure. I stand by everything I wrote.

  5. Extracting more than 15 to 20% of wáter from ANY Reverse Osmosis Membrane is going to plug it prematurely most times, and the reason why home owners have to replace membranes on a regular basis (Good for vendors!). While it is possible that low mineral content municipal water in some places could mean a lower reject can be set, most places will have water that will plug the membrane eventually when attempting to get more water and less reject. Unless you are installing a home RO system in a place with very scarce wáter, a biger reject is beneficial for the membrane, and usually for the RO owner too.

    A larger (SEVERAL times larger) UV unit then is required, and it is going to cost more. This needs to be announced to the prospective buyer before stating it as mandatory.

    Plastic tubing degradation and failure is more closely related to poor quality Chinese tubing (Although very aceptable quality Chinese made ones are available, only more difficult to get, but available). Poor quality tubing degradation is going to happen either at the beginning or at the end of the system, and the higher pressure at entrance could make it fail sooner and produce a bigger flooding. UV resistance of plastics is dependent on UV pigments and stabilizers, as well as base resin type and quality. Many plastic tubing manufacturers, not only from China, use recycled polymers, some even claiming “Green” reasons, but production cost is the real reason. The warm temperaturas that UV lamps in Domestic RO systems do NOT cause embrittlement at all (please study something called “Glass Transition Temperature”). Embrittlement is caused by Ozone, UV rays and por quality polymers loosing plastizicers (of low quality too).

    Lastly, I woukld like to know how effective has to be an UV sterilizer that has to treat all the wáter used in a home, like power, UV intensity, contact time etc. One thing you apparently forget to mention and consider is the effect of wáter contaminants on the effectiveness of the UV lamp itself: the blocking of UV light by the contaminants themselves! (another reason the UV is not best put at the Point of Entrance…)

    I’ll repeat the paragraph of a UV lamp manufacturer that will put it in an appropriate manner:

    “Selecting your UV System: The quality or appropriateness of both the UV light and of the ‘contact ‘, are crucial to accomplish disinfection. It is important to properly ‘size’ the UV based upon the application. The contact disinfectant property of UV light energy is measured in microwatt-seconds per square centimeter (uw-sec/cm2). LOW PRESSURE lamps which are best suited for small intermittent flow applications. When selecting a system be aware of your use requirements (gallons per day) and your preferred flow rates. It is equally important to use a good pre-filter to remove any dirt or debris that may be present in the raw water supply. This dirt and debris can interfere with the effectiveness of the UV rays – virtually giving the microorganism a shield to protect them when passing the UV rays.”

    Thus, depending on the particular situation, the UV will be more convenient to be installed either after or before the RO. If your wáter is so heavily contaminated as to require a drastic UV sterilization for ALL uses in the home, then suggest to put it before the RO, but with the proper filtration. But in many cases of municipal wáter, putting the UV after the RO is more sensible.

    More than continuing a fight, I am trying to provide some considerations that I did not saw in your brief article. MAybe readers deserve a better and more balanced explanation on the merits against the disadvantages of putting the UV before or after the RO system. In that manner, all will be enlightened. Best luck.

  6. It’s clear to me that you are writing this from a theoretical viewpoint from a 20 year-old book you might have read. The fact of the matter that in the past few years membrane technology has improved to the point to where we routinely achieve 75% recovery with 4 x 40 membranes… and they last for years. In residential applications we routinely have 25 to 33% recovery on membranes which last about two or three years. A residential 50 GPD RO membrane costs only about $34.00, so it’s not a big deal anymore. We have 75% recovery systems where the membranes last 3+ years. The secret is proper pre-treatment with proper filtration, absorption, softening or an ant-scalant and even UV when microbiological fouling is present.

    I am not talking about cheap tubing or UV Degradation – I am talking about a small chamber where a UV bulb can heat the water to over 170 degrees and thermal expansion and contraction. In 40+ years and tens of thousands of installations, I have seen horrible disasters, so much so that I advise people not to do point-of-use UV, but rather to do the whole house for many reasons already mentioned. We suggest sizing the UV systems to a flow rate of @40mJ/cm2. Yes, it’s more expensive, but it protects everyone throughout the home and is a lot cheaper and less inconvenient that a flood in your kitchen.

    It is understood that UV requires proper pre-treatment and we preach that everyday. We have many articles and other information on this and each article is not intended to be an end-all, be all treatise, but rather just an aspect of certain water treatment practices.

  7. Well Mr. Timmons, I am a customer that stumbled across the blog while trying to determine where to locate a UV tube on my RO system. An impartial reader that was considering both the opinions presented. You loose the argument primarily due to your insulting responses. Secondly due to you’re resistance to understand the valid parts of an opposing point of view. And lastly by using an infection contracted in a third world country as a reason people should waste a lot of money and electricity here in the U.S. A 50 watt to 90 watt whole house UV uses 5 times the electricity (about $70 per year compared to $10) as a 10-14 watt unit. A yearly replacement bulb adds another $100 + to the annual cost. And for what? To water your lawn, wash your car, flush the toilet, wash your clothes…… Less than 1 person in a million in the US suffers from a cutaneous infection with tap or city water per year. I know people that use UV on their entire water supply. They are very wealthy and do not care about a few hundred dollars every year. They also need to disinfect their entire home plumbing including the RO every time they do any minor plumbing work because contamination can occur AFTER the UV. To sum up your refusal to see the impractical logic in your argument does not validate you expertise. Your readers are not as uniformed as you continue to insist. They see your point of few and understand your thinking. Do you have that same ability? You are loosing credibility.

  8. Steve,

    If my responses are insulting, then you are the master of insults. You are very hypocritical! I merely stated that installing a point of use ultraviolet disinfection system was not a very good way of using ultraviolet. If you disagree with that you are very uninformed! Evidently, you have never seen a pipe burst or a fitting melt from the heat generated by the UV bulb in a small confined area. Maybe you have never seen a person get cholera or eye infections from showering in contaminated water. My point is: IF YOU HAVE CONTAMINATED WATER YOU SHOULD TREAT THE ENTIRE HOUSE IF POSSIBLE! What is your problem? Do you sell undersink UV’s and have a financial interest in this?

    Listen, have you ever seen a little kid drink out of a hose? What does it hurt to kill the bacteria? You have no argument – it’s obvious you have some agenda – I don’t know what it is, but it’s devoid of reason! You are also wrong about the cost – you have way overstated it!

    No, my readers are not uninformed, but YOU ARE! Look, I am talking about a device that kills bacteria and saves lives and you are complaining about that? Geeeezzzzz! I’ll block you next time – I just want people to see how silly your argument is.

  9. I used to have an undersink unit. I made the mistake of using the plastic tubing which did rupture over time. I do how ever feel compelled to try again. Would it be possible to run stainless steel lines at least 15 to 20 inches long on the inlet and outlet to the ro faucet so as not to degrade any other plastic parts?

  10. We have two new products, both of which will be released between now and June for undersink drinking water systems.

    One is called the Silecte inline filter and the other is the Pulsar inline filter. Both kill bacteria better than UV! Watch for them. Stainless steel lines? Not a good idea in my opinion…

  11. Hello,
    Why do you say to install the uv light at point of entry?
    I have very hard water, 42 grains. These are very large particles that the uv light may not be able to penetrate. Would it not make more sense to install it after all other treatments/filters, such a water softener etc?

  12. Thanks for your very informative discussions.
    I am planning on starting a water bottling filling station with an RO system and UV sterilization . The source of the water is from a well in Ontario, Canada. The water hardness is 7, the iron(ppm) .08, the Ph 7.5, TDS 280. Most of my customers will be in a country setting on wells. I will be processing the water in advance of distribution. Storing the RO water in large holding tanks and then distributing it from the holding tanks. Question Should I put the UV before the RO system. I am looking at a Vectapure RSX Lite 300 GPD RO system.
    Thanks Vince.

  13. Vince,

    Are you planning on not selling much water? The reason I ask is because that is a very light-duty system and with your water temperature, it will make about 150 gallons per day (at best). Is that enough? How big are your holding tanks? I would not use UV – I would use Quantum Disinfection. The disinfection should be the very last thing, although if your water is microbiologically unsafe, you may want to do both. You can contact me at mark@uswatersystems.com

  14. Hey Mark,

    I’m not sure about UV being last. I treat rainwater and had particulate -> carbon block -> UV

    Using Poly particulate filters so bacteria wouldn’t grow, but the carbon block would clog from growth.

    Calling Pentek they told me I did it wrong. It should be particulate filter (I have two, last stage is 1 micron) then UV then into carbon block.

    Sure enough I redid system and my 0.5 micron 4.5×20 carbon block lasts 3x as long now.

Leave a Reply

Your email address will not be published. Required fields are marked *